Much research focuses on what might be possible with digital games in the classroom. This study focuses on what is currently probable and typical. It uses a controlled quasi-experimental design to compare outcomes for students of 13 teachers in 10 diverse urban, suburban, and rural schools. The teachers integrated a set of 55 typical educational games into their curricula on Jacksonian democracy. Teachers reported strong engagement benefits for the game condition and a strong interest in using games of this type in the future in their surveys and interviews. Each teacher taught at least one classroom with the games and at least one classroom without the games. When the one teacher who reported a failed implementation was dropped from the analysis, the results showed significantly higher gains for the game condition in terms of multiple-choice factual outcomes, open-response factual outcomes, evidentiary depth, and student engagement outcomes. When the failed implementation was included, the game condition demonstrated improved scores for all outcomes, but only student engagement and evidentiary depth remained significant. Moderator analyses highlighted the role of teacher experience and student engagement in the efficacy of the game condition, indicating that game instruction was particularly beneficial for special education students.